Recently, the increasing frequency of torrential rainfall due to climate change has heightened the risk of embankment and slope failures, leading to a growing need for slope stability monitoring. In particular, levees are critical structures that requ...
Recently, the increasing frequency of torrential rainfall due to climate change has heightened the risk of embankment and slope failures, leading to a growing need for slope stability monitoring. In particular, levees are critical structures that require ensuring hydraulic stability, making precise displacement observation essential. Radar interferometry based on ground-based radar is effectively utilized for monitoring geological phenomena, such as land subsidence and landslides, as well as the stability of man-made structures like dams and levees, with millimeter-level precision. Ground-based radar allows for flexible configuration of the image acquisition area and time based on the user's objectives, facilitating the acquisition of high-precision time-series data due to its high temporal and spatial resolution. However, the atmospheric phase delay effect, which occurs as microwaves propagate through the troposphere, acts as a major source of error in precise displacement observations. In this study, atmospheric phase correction based on meteorological observation data and the Persistent Scatterer Interferometry(PSI) was applied to ground-based radar data acquired at the Wangsin reservoir levee in Gyeongju, Gyeongsangbuk-do. We evaluated the interferometric phase and time-series displacement before and after atmospheric correction. The Ku-band Gamma Portable Radar Interferometer (GPRI-Ⅱ) was used to acquire 397 radar images at 5-minute intervals, along with simultaneous meteorological data, from 17:00 on September 18, 2025, to 02:00 on September 20, 2025. The observation period was divided into 11 time segments to generate an initial atmospheric phase model using atmospheric refractivity, which was then refined using the actual observed interferometric phase. The results showed that the residual phase in the differential interferogram was corrected to a value close to zero after atmospheric correction. Additionally, the linear error displacement velocity induced by atmospheric effects was also corrected to nearly 0 mm/day. By selecting four specific persistent scatterer points and examining their time-series displacements, a significant reduction in the Root Mean Square Error(RMSE) to near-zero values was confirmed, validating the successful performance of the atmospheric correction. Based on these results, the time segment from 20:05 to 23:00, which exhibited the highest correlation between atmospheric refractivity and interferometric phase along with successful correction performance, was identified as the optimal observation window. This period is considered the optimal time for observations when applying atmospheric phase correction based on meteorological data. This study constructed the atmospheric phase model assuming a uniform atmospheric refractivity across the entire area based on a single point. Although this approach has a limitation in that it does not account for localized meteorological changes or wind effects, it successfully corrected the overall atmospheric phase in the ground-based radar data. These findings evaluate the persistent scatterer interferometric phase through meteorological-based atmospheric correction and suggest the possibility of more precise time-series displacement observation.